396 



Mr. A, LI. Hughes on the 



To obtain the distribution of velocities of the electrons, the 

 leak is measured with the case D at different potentials 

 ranging from +40 volts to —8 volts. The arc is struck, 

 and in a few seconds after which the current through it has 

 become approximately steady, the earth connexion of the 

 electrometer is broken and the leak is measured over an 

 interval of time of a minute. It was not safe to run the arc 

 for much longer as the heating appeared to produce a little 

 gas, making the readings somewhat unsteady. 



Readings were also taken with the quartz plate Q inter- 

 posed in the path of the light. The following are the 

 results obtained for the leaks from the illuminated plate B 

 with and without the quartz in the path of the light. 



Table I. 

 Source of Light — Mercury Arc. 



Potential 

 of D. 



Leak without quartz. 



Leak with 

 quartz. 



+40 volts 



-H90( = l-06xl0- n 



4-226 



•+■ 4 



472 amp.) 



225 



+ 2 



482 



217 



+ 1 



412 



202 







305 



127 



- -34 



195 



GO 



- -66 



127 



21 



-1-00 



67 



9 



-134 



34 



+ 1-3 



-1-66 



19 



-1-3 



-2-00 



12 



-2-0 



-2-34 



. 9 





-2-66 



+ 5 





-3-00 







— 2 - 2 



-3-34 



- 3 





-3-36 



- 5 





-4-00 



- 7 





-6-00 



- 9 



-2-7 



-8-00 



-10 





To determine the potential at which the leak was zero, the 

 capacity was cut out of the electrometer system and the sign 

 of the leak was observed for different potentials of D. When 

 the potential was —3*0 volts or less the leak was always 

 positive, when it was — 3'1 or more it was always negative. 

 With potentials in between these limits the sign of the leak 

 for a given potential was sometimes negative and sometimes 

 positive, and therefore the potential for zero leak could only 

 be indicated by the above' limits. 



